HiLo — How to play & pay table

Everything about the game: how it plays, what each combination pays, the RTP, and how every round can be verified on-chain.

🃏

HiLo

Higher or lower — keep the winnings, decide again

House Edge: 1% – 1.54%

HiLo is the classic higher-or-lower card game, running entirely on-chain — with one honest difference: there is no hand. You hold a rank, call the next one higher or lower, and each call is its own independent bet. Win and it pays into your Balance immediately, the drawn rank becomes the card you hold, and the next call re-stakes exactly that amount — so the streak you build is real money changing hands at every step.

🛡️

Fairness

There is no deck to trust, because there is no deck. The chain produces verifiable randomness (Chainlink VRF on Arbitrum, Switchboard on Solana — committed before your bet and verified at reveal), and each draw is ONE division of it: rank = randomness mod 13 + 1. Take the published value, run the same division, and you get the same rank. The price of a seat is ⌊99·13/ways⌋ percent, where ways is how many of the thirteen ranks win it — the shaded run on the strip is literally that numerator — from a table generated from a single source and compiled into the program. That floor is the honest part: a call has one payout knob and the label grid is 0.01×, so the return lands at 99% or just under it, never above. Seats therefore differ a little — 98.46% to 99.00%, a house edge of 1.00% to 1.54% — and each button prints its own figure.

Randomness by: Chainlink VRF · SwitchboardContract operated by: DODEN.IO

Frequently asked questions

How does HiLo work?: You hold a card (Ace through King) and call whether the next rank lands HIGHER or LOWER. EVERY CALL IS ITS OWN BET: call it right and the winnings land in your Balance, the drawn rank becomes the card you now hold, and the next call automatically stakes exactly what you just won. Call it wrong and that stake is gone and the run ends.
What happens on a tie?: A tie LOSES. The program pays only on a strictly higher rank (HI) or a strictly lower one (LO) — and the price already accounts for that: HI on card c wins on 13 − c ranks and LO wins on c − 1, which is exactly the shaded run on the 13-rank strip under the table.
Why is one of the buttons locked?: Two seats are impossible: HI on a King and LO on an Ace have zero ranks that could win. The table holds a 0 there and the program refuses them outright (`choice_value_ok`), so the interface disables the button with its reason and never sends the bet — taking it would not be a wager, it would be a donation.
Who picks the first card?: Your browser draws it, and you can redraw it as often as you like before a run starts. That is safe because EVERY (direction, card) pair is priced by the SAME published rule: the chain holds no deck, it prices the seat you claim to be sitting in, so no redraw can produce a seat the table has not already committed to. To be precise, though: the seats do NOT all return the same. Rounding down onto the 0.01× label grid lands them between 98.46% and 99.00%, so the seat you hold does shift your return a little. That is why each HI/LO button prints its own RTP — you can see which seat you are taking before you take it. And there are no suits anywhere, because the program publishes only a rank.
Is "cash out" a transaction?: No. There is no multi-step round on-chain: each call has already settled and already paid into your Balance. Stopping is simply not calling again. Nothing to sign, nothing that can fail, and no money can ever be stranded in a half-finished round.
Where do the multipliers come from?: Pure arithmetic: ⌊99·13/ways⌋ percent, where ways is how many ranks win that seat. Holding an Ace and calling HI pays 1.07× (12 of 13 ranks win); holding a Queen and calling HI pays 12.87× (only a King wins). The table is generated from a single source, lives in the on-chain program, and every entry sits on the 0.01× grid.
So is the RTP exactly 99%?: No, and we do not claim it is. A call has exactly ONE payout knob and the label grid is 0.01×, so rounding DOWN always lands the return at 99% or below — never above. Where it lands depends on how cleanly 99·13 divides by the number of winning ranks, so the real return runs from 98.46% to 99.00% (a house edge of 1.00%–1.54%). Only 8 of the 24 playable seats land on 99.00% exactly (HI on a Queen and HI on a 10, among others); the thinnest is a FOUR-WAY TIE on RETURN: HI on a 3 and LO on a J at 1.28×, HI on a 5 and LO on a 9 at 1.60× — different multipliers, the same 98.46%. An Ace called HI returns 98.77%. This is exactly how Stake and BC.game build their hilo tables. Because the figure changes seat by seat, each button prints its own.